Soft pneumatic actuators (SPAs) are extensively investigated due to their simple control strategies for producing sophisticated motions. However, the motions or operations of homogeneous SPAs show obvious limitations in some varying curvature interaction scenarios because of the profile mismatch of homogeneous SPAs and specific interacted objects. Herein, a stiffness preprogrammable soft pneumatic actuator (SPSPA) is proposed by discretely presetting gradient geometrical or materials distributions. Through finite element analysis and experimental validation, a mathematical model of behavior prediction of SPSPA was built to relate the geometrical parameters/materials with its morphing behaviors, making it possible to reversely obtain designed parameters. This design strategy enables conformal and efficient interaction in some curvature varying scenarios. Specifically, higher effective contact area, perimeter utilization ratio, and conformal ability can be obtained while interacting with those inhomogeneous curvature objects, for example, more than 434.7% improvement in contact area rates and 12.5% enhancement in perimeter utilization ratios toward a typical equilateral triangle object. Further, a serial of SPSPAs that have conformal grasping/interactive capability, better contact sensing behaviors were demonstrated. For example, an SPSPA and an SPSP robot were demonstrated, which showed better kinetic, kinematic characterizations and sensing capability compared with the homogeneous one while coming across varying curvature objects. Moreover, underactuated finger rehabilitation SPSPAs were demonstrated with customized profiles and coupled joint motion. This customized scheme can be potentially used in those specific-purposed, single, and repetitive application scenarios where varying curvature, conformal and efficient interaction are needed.

中文翻译：

---

用于高效、保形操作的刚度预编程软弯曲气动执行器

软气动执行器 (SPA) 因其用于产生复杂运动的简单控制策略而受到广泛研究。然而，由于同质 SPA 和特定交互对象的轮廓不匹配，同质 SPA 的运动或操作在一些不同的曲率交互场景中显示出明显的局限性。在此，通过离散预设梯度几何或材料分布提出了一种刚度可预编程软气动致动器（SPSPA）。通过有限元分析和实验验证，建立了SPSPA行为预测的数学模型，将几何参数/材料与其变形行为联系起来，使反向获得设计参数成为可能。这种设计策略可以在一些曲率变化的场景中实现保形和高效的交互。具体而言，在与那些曲率不均匀的物体相互作用时，可以获得更高的有效接触面积、周长利用率和保形能力，例如，接触面积率提高 434.7% 以上，周长利用率提高 12.5% 对典型的等边三角形目的。此外，还展示了一系列具有保形抓取/交互能力、更好的接触传感行为的 SPSPA。例如，展示了一个 SPSPA 和一个 SPSP 机器人，与同质机器人相比，它们在遇到不同曲率的物体时表现出更好的动力学、运动学特征和传感能力。而且，欠驱动的手指康复 SPSPA 用定制的轮廓和耦合的关节运动进行了演示。这种定制方案可以潜在地用于那些需要不同曲率、保形和高效交互的特定用途、单一和重复的应用场景。